Author: tlinnet
Date: Wed Aug 27 08:50:56 2014
New Revision: 25321
URL: http://svn.gna.org/viewcvs/relax?rev=25321&view=rev
Log:
Modified profiling script, but it seems that the dfunc from
target_functions.relax_fit does not work.
task #7822(https://gna.org/task/index.php?7822): Implement user function to
estimate R2eff and associated errors for exponential curve fitting.
Modified:
trunk/test_suite/shared_data/curve_fitting/profiling/profiling_relax_fit.py
Modified:
trunk/test_suite/shared_data/curve_fitting/profiling/profiling_relax_fit.py
URL:
http://svn.gna.org/viewcvs/relax/trunk/test_suite/shared_data/curve_fitting/profiling/profiling_relax_fit.py?rev=25321&r1=25320&r2=25321&view=diff
==============================================================================
---
trunk/test_suite/shared_data/curve_fitting/profiling/profiling_relax_fit.py
(original)
+++
trunk/test_suite/shared_data/curve_fitting/profiling/profiling_relax_fit.py
Wed Aug 27 08:50:56 2014
@@ -60,6 +60,18 @@
from target_functions.relax_fit import setup, func, dfunc, d2func,
back_calc_I
from target_functions.chi2 import chi2_rankN
+## Set the min_algor.
+## simplex is algorithms without gradient. It is quite slow, since it needs
to take many steps.
+#min_algor='simplex'
+
+## Steepest descent uses only the gradient. This works, but it is not
totally precise.
+##min_algor = 'Steepest descent'
+##max_iterations = 1000
+
+## Quasi-Newton BFGS. Uses only the gradient.
+## This gets the same results as 2000 Monte-Carlo with simplex.
+## This is one of the best optimisation techniques when only the function
and gradient are present, as it tries to numerically approximate the Hessian
matrix, updating it as the algorithm moves along.
+#min_algor = 'BFGS'
# Alter setup.
def main():
@@ -75,6 +87,9 @@
sum_root_squared = sum( sqrt( (v_cT_chi2_list - v_cF_chi2_list)**2 )
)
print("The sum of the root squared differences between with and
without constraints are: %.3e" % sum_root_squared)
+ # Calculate without contraints, BFGS.
+ #v_BFGS_cF_chi2_list = array(verify(min_algor='BFGS',
constraints=False))
+
# Now calculate with Python code.
# Calculate with contraints.
v_pyt_cT_chi2_list = array(verify_pyt(constraints=True))
@@ -99,7 +114,7 @@
print("C_cT=%1.1e C_cF=%1.1e P_cT=%1.1e P_cF=%1.1e" %
(chi_v_cT-chi_v_cT, chi_v_cF-chi_v_cT, chi_v_pyt_cT-chi_v_cT,
chi_v_pyt_cF-chi_v_cT) )
# Do verification for Python code, and difference between minfx and
Scipy optimisation without constraints.
- if False:
+ if True:
# Calculate with Python code.
# Calculate without contraints.
v_pyt_cF_chi2_list = array(verify_pyt(constraints=False))
@@ -118,9 +133,7 @@
# Do profiling.
- lines_report = 15
-
- if False:
+ if True:
#################
print("Verify, with constraints, C code")
constraints = True
@@ -129,22 +142,41 @@
verbose = True
# Calc for verify with constraints.
- v_cT_filename = tempfile.NamedTemporaryFile(delete=False).name
- # Profile for a single spin.
- cProfile.run('verify(constraints=%s)'%constraints, v_cT_filename)
-
- # Read all stats files into a single object
- v_cT_stats = pstats.Stats(v_cT_filename)
-
- # Clean up filenames for the report
- v_cT_stats.strip_dirs()
-
- # Sort the statistics by the cumulative time spent in the function.
cumulative, time, calls
- v_cT_stats.sort_stats('cumulative')
-
- # Print report for single.
- if verbose:
- v_cT_stats.print_stats(lines_report)
+ filename = tempfile.NamedTemporaryFile(delete=False).name
+ # Profile for a single spin.
+ cProfile.run('verify(constraints=%s)'%constraints, filename)
+
+ print_report(filename=filename, verbose=verbose)
+
+ if True:
+ #################
+ print("Verify, without constraints, C code")
+ constraints = False
+
+ # Print statistics.
+ verbose = True
+
+ # Calc for verify with constraints.
+ filename = tempfile.NamedTemporaryFile(delete=False).name
+ # Profile for a single spin.
+ cProfile.run('verify(constraints=%s)'%constraints, filename)
+
+ print_report(filename=filename, verbose=verbose)
+
+ if True:
+ #################
+ print("Verify, without constraints, C code BFGS")
+ constraints = False
+
+ # Print statistics.
+ verbose = True
+
+ # Calc for verify with constraints.
+ filename = tempfile.NamedTemporaryFile(delete=False).name
+ # Profile for a single spin.
+ cProfile.run('verify(min_algor="BFGS",
constraints=%s)'%(constraints), filename)
+
+ print_report(filename=filename, verbose=verbose)
if False:
#################
@@ -155,48 +187,11 @@
verbose = True
# Calc for verify with constraints.
- v_pyt_cT_filename = tempfile.NamedTemporaryFile(delete=False).name
- # Profile for a single spin.
- cProfile.run('verify_pyt(constraints=%s)'%constraints,
v_pyt_cT_filename)
-
- # Read all stats files into a single object
- v_pyt_cT_stats = pstats.Stats(v_pyt_cT_filename)
-
- # Clean up filenames for the report
- v_pyt_cT_stats.strip_dirs()
-
- # Sort the statistics by the cumulative time spent in the function.
cumulative, time, calls
- v_pyt_cT_stats.sort_stats('cumulative')
-
- # Print report for single.
- if verbose:
- v_pyt_cT_stats.print_stats(lines_report)
-
- if True:
- #################
- print("Verify, without constraints, C code")
- constraints = False
-
- # Print statistics.
- verbose = True
-
- # Calc for verify with constraints.
- v_cF_filename = tempfile.NamedTemporaryFile(delete=False).name
- # Profile for a single spin.
- cProfile.run('verify(constraints=%s)'%constraints, v_cF_filename)
-
- # Read all stats files into a single object
- v_cF_stats = pstats.Stats(v_cF_filename)
-
- # Clean up filenames for the report
- v_cF_stats.strip_dirs()
-
- # Sort the statistics by the cumulative time spent in the function.
cumulative, time, calls
- v_cF_stats.sort_stats('cumulative')
-
- # Print report for single.
- if verbose:
- v_cF_stats.print_stats(lines_report)
+ filename = tempfile.NamedTemporaryFile(delete=False).name
+ # Profile for a single spin.
+ cProfile.run('verify_pyt(constraints=%s)'%constraints, filename)
+
+ print_report(filename=filename, verbose=verbose)
if False:
#################
@@ -207,25 +202,13 @@
verbose = True
# Calc for verify with constraints.
- v_pyt_cF_filename = tempfile.NamedTemporaryFile(delete=False).name
- # Profile for a single spin.
- cProfile.run('verify_pyt(constraints=%s)'%constraints,
v_pyt_cF_filename)
-
- # Read all stats files into a single object
- v_pyt_cF_stats = pstats.Stats(v_pyt_cF_filename)
-
- # Clean up filenames for the report
- v_pyt_cF_stats.strip_dirs()
-
- # Sort the statistics by the cumulative time spent in the function.
cumulative, time, calls
- v_pyt_cF_stats.sort_stats('cumulative')
-
- # Print report for single.
- if verbose:
- v_pyt_cF_stats.print_stats(lines_report)
-
-
- if False:
+ filename = tempfile.NamedTemporaryFile(delete=False).name
+ # Profile for a single spin.
+ cProfile.run('verify_pyt(constraints=%s)'%constraints, filename)
+
+ print_report(filename=filename, verbose=verbose)
+
+ if True:
#################
print("Verify, without constraints, Python Scipy")
@@ -233,22 +216,28 @@
verbose = True
# Calc for verify with constraints.
- v_sci_cF_filename = tempfile.NamedTemporaryFile(delete=False).name
- # Profile for a single spin.
- cProfile.run('verify_sci()', v_sci_cF_filename)
-
- # Read all stats files into a single object
- v_sci_cF_stats = pstats.Stats(v_sci_cF_filename)
-
- # Clean up filenames for the report
- v_sci_cF_stats.strip_dirs()
-
- # Sort the statistics by the cumulative time spent in the function.
cumulative, time, calls
- v_sci_cF_stats.sort_stats('cumulative')
-
- # Print report for single.
- if verbose:
- v_sci_cF_stats.print_stats(lines_report)
+ filename = tempfile.NamedTemporaryFile(delete=False).name
+ # Profile for a single spin.
+ cProfile.run('verify_sci()', filename)
+
+ print_report(filename=filename, verbose=verbose)
+
+
+def print_report(filename=None, verbose=True):
+ lines_report = 1
+
+ # Read all stats files into a single object
+ stats = pstats.Stats(filename)
+
+ # Clean up filenames for the report
+ stats.strip_dirs()
+
+ # Sort the statistics by the cumulative time spent in the function.
cumulative, time, calls
+ stats.sort_stats('cumulative')
+
+ # Print report for single.
+ if verbose:
+ stats.print_stats(lines_report)
class Profile:
@@ -308,7 +297,7 @@
yield values[:num_times], errors[:num_times],
times[:num_times], struct[:num_times], num_times
-def verify(constraints=None):
+def verify(min_algor='simplex', constraints=None):
# Instantiate class.
C = Profile()
@@ -325,19 +314,6 @@
# Initial guess for minimisation. Solved by linear least squares.
x0 = asarray( E.estimate_x0_exp() )
- # Set the min_algor.
- # simplex is algorithms without gradient. It is quite slow, since it
needs to take many steps.
- min_algor='simplex'
-
- # Steepest descent uses only the gradient. This works, but it is not
totally precise.
- #min_algor = 'Steepest descent'
- #max_iterations = 1000
-
- # Quasi-Newton BFGS. Uses only the gradient.
- # This gets the same results as 2000 Monte-Carlo with simplex.
- # This is one of the best optimisation techniques when only the
function and gradient are present, as it tries to numerically approximate the
Hessian matrix, updating it as the algorithm moves along.
- min_algor = 'BFGS'
-
E.set_settings_minfx(min_algor=min_algor, constraints=constraints)
# Initialise the function to minimise.
@@ -355,7 +331,7 @@
return chi2_list
-def verify_pyt(constraints=None):
+def verify_pyt(min_algor='simplex', constraints=None):
# Instantiate class.
C = Profile()
@@ -372,19 +348,6 @@
# Initial guess for minimisation. Solved by linear least squares.
x0 = asarray( E.estimate_x0_exp() )
- # Set the min_algor.
- # simplex is algorithms without gradient. It is quite slow, since it
needs to take many steps.
- min_algor='simplex'
-
- # Steepest descent uses only the gradient. This works, but it is not
totally precise.
- #min_algor = 'Steepest descent'
- #max_iterations = 1000
-
- # Quasi-Newton BFGS. Uses only the gradient.
- # This gets the same results as 2000 Monte-Carlo with simplex.
- # This is one of the best optimisation techniques when only the
function and gradient are present, as it tries to numerically approximate the
Hessian matrix, updating it as the algorithm moves along.
- #min_algor = 'BFGS'
-
E.set_settings_minfx(min_algor=min_algor, constraints=constraints)
# Define func.
@@ -421,7 +384,7 @@
# Unpack results
param_vector, param_vector_error, chi2, iter_count, f_count,
g_count, h_count, warning = results
- if E.verbosity:
+ if print_info:
r2eff = param_vector[0]
i0 = param_vector[1]
r2eff_err = param_vector_error[0]
r25321 - /trunk/test_suite/shared_data/curve_fitting/profiling/profiling_relax_fit.py